Local Law 84 benchmarking looks like the easiest filing in the NYC compliance calendar. You log in to Energy Star Portfolio Manager, enter your utility bills, generate a report, and submit. In practice, owners get this wrong constantly — and because LL84 data feeds LL97 emissions reporting, getting it wrong twice now compounds into real money. Here is what to watch for.
A quick refresher on LL84
LL84 requires owners of buildings over 25,000 square feet to benchmark whole-building energy and water use annually using EPA's Energy Star Portfolio Manager. The benchmarking submission is due May 1 each year and covers the prior calendar year.
The penalty for not filing is $500 per quarter — capped at $2,000 per year. That sounds small, but the data flows directly into your LL97 report and into the public dataset DOE publishes, which lenders and insurers increasingly check during diligence. A messy LL84 record creates downstream problems that cost much more than the fines.
Mistake 1: Stale or incomplete property setup
Portfolio Manager properties drift. A building gets renovated, occupancy changes, a parking garage gets converted to retail — and the property setup in PM never reflects it. The benchmarking score generated from outdated property attributes is meaningless, and worse, the LL97 cap derived from those occupancy types will be wrong.
Before you enter a single bill, audit the property setup:
- Gross floor area. Does it match your CO and your tax records? PM uses gross, not net.
- Property type and use details. A multifamily property in PM has different required attributes (number of units, bedrooms, common space heating) than an office. Walk through them.
- Year built and significant renovations. Affects the benchmark comparison.
- Number of weekly operating hours. For commercial buildings this drives the score.
Mistake 2: Whole-building data, but only the master meter
LL84 requires whole-building energy use. If the master meter only covers common areas and tenants are on direct meters, you are reporting a fraction of the actual consumption. The score will look fantastic and DOE will know it is wrong.
For Con Edison electric and gas, request Aggregated Whole-Building Data through the Con Ed Demand Portal. They can deliver monthly totals by fuel for the whole building once you have four or more covered tenants. National Grid has a similar program.
For steam, oil, and propane, you typically have full-building visibility because there is one delivery to the building. But verify — some properties have separate steam taps for ground-floor retail.
If you cannot get aggregated tenant data because you have fewer than four covered tenants, the rule allows you to report what you have and note the limitation. Do that explicitly. Don't fabricate, don't extrapolate, and don't pretend the master meter is the whole building.
Mistake 3: Missing fuels
A building has every fuel it has. Buildings powered by electricity, natural gas, and steam need three meter setups in PM. Buildings with backup generators that run on diesel need that fuel reported too. Buildings with on-site solar need the solar production entered as a separate meter.
Owners frequently forget:
- District steam. If you receive Con Ed steam, that is its own meter type in PM and its own carbon coefficient in LL97.
- Fuel oil. If you have a dual-fuel boiler that runs on oil during peak gas curtailment, you need oil delivery records too. The fact that it ran for two weeks does not exempt it from reporting.
- Solar generation. Behind-the-meter solar reduces grid purchases but the gross generation has to be entered as well.
Missing a fuel doesn't lower your emissions on the LL97 side. DOB cross-references benchmarking against utility records during audits and you will be flagged.
Mistake 4: Vacancy and irregular operations not noted
The pandemic taught everyone that vacancy assumptions matter. PM lets you note when a building had unusual occupancy — full lockdown, partial reopening, major renovation — and it factors into the score generation.
If your building had 20% occupancy in 2024 because it was undergoing renovation, the energy use intensity will look misleadingly low or high depending on how the unoccupied floors were heated. Record it. The score is comparative; without context, your future scores will look like regressions when they are actually a return to baseline.
Mistake 5: Submitting without a confirmation
The submission step is where filings get lost. Generating the report in PM is not the same as submitting it to NYC. The submission goes through PM's "Submit Data" workflow under the LL84 reporting template, and it returns a confirmation number. Without that confirmation number, you have not filed.
Save the confirmation. Save it in two places. Email it to yourself. The single most common LL84 violation is not "the building is over its energy budget" — it is "the owner thought they filed and didn't."
A workflow that actually works
If you are responsible for benchmarking three or more buildings, treat this like a small project with a clear sequence:
- January: Pull the previous year's utility bills. Reconcile them against PM. Resolve missing months now before the data window closes.
- February: Audit property attributes. Update gross floor area, occupancy, hours.
- March: Request Con Ed aggregated data. It can take 4 to 6 weeks. Don't wait until April.
- Early April: Run the benchmarking report. Review for outliers — a 40% jump in consumption usually means a missing meter, not a real change.
- Mid April: Submit through PM. Save the confirmation.
- Post submission: Use the data to populate your LL97 emissions calculation. The numbers should match exactly.
If the LL84 number and the LL97 number for the same building, same year, don't match — one of them is wrong. Reconcile before May 1.
What an audit looks like
DOE audits a fraction of submissions every year. They request:
- Twelve months of utility bills, by fuel.
- The PM property setup documentation.
- Any aggregated data requests and Con Ed responses.
- A description of how vacant floors were treated.
If the data lines up cleanly, the audit closes in a few weeks. If it doesn't, you get a notice of violation and a request to amend. Amendments are not subtle — they show up in the public dataset with a flag.
For owners with multi-building portfolios, this is exactly the kind of work that lives better inside a tracking system than a spreadsheet. LLDesk keeps the benchmarking deadline, the property setup, and the LL97 emissions data in sync per building so the May 1 numbers reconcile by default.
Key takeaways
- LL84 benchmarking is due May 1 for every NYC building over 25,000 sf. The penalty is $500 per quarter, capped at $2,000 per year.
- The big errors are stale property setup, missing tenant data, missed fuels, ignored vacancy, and submissions that were never actually submitted.
- Request Con Ed aggregated whole-building data in March, not April. It takes weeks.
- Reconcile your LL84 inputs against your LL97 emissions report before submitting either.
- The public dataset is consulted by lenders and insurers during diligence. Get it right the first time.